from abc import ABC, abstractmethod
from typing import Union
import xarray as xr
import numpy as np

class BaseCopApproximator(ABC):
    """
    Abstract class for approximating the coefficient of performance (COP) of a heat pump."""
    def __init__(
            self,
            forward_temperature_celsius: Union[xr.DataArray, np.array],
            source_inlet_temperature_celsius: Union[xr.DataArray, np.array],
        ):
            """
            Initialize CopApproximator.

            Parameters:
            ----------
            forward_temperature_celsius : Union[xr.DataArray, np.array]
                The forward temperature in Celsius.
            return_temperature_celsius : Union[xr.DataArray, np.array]
                The return temperature in Celsius.
            """
            pass
 
    @abstractmethod
    def approximate_cop(self) -> Union[xr.DataArray, np.array]:
        """Approximate heat pump coefficient of performance (COP).

        Returns:
        -------
        Union[xr.DataArray, np.array]
            The calculated COP values.
        """
        pass
 
    def celsius_to_kelvin(t_celsius: Union[float, xr.DataArray, np.array]) -> Union[float, xr.DataArray, np.array]:
        if (np.asarray(t_celsius) > 200).any():
            raise ValueError("t_celsius > 200. Are you sure you are using the right units?")
        return t_celsius + 273.15


    def logarithmic_mean(t_hot: Union[float, xr.DataArray, np.ndarray], t_cold: Union[float, xr.DataArray, np.ndarray]) -> Union[float, xr.DataArray, np.ndarray]:
        if (np.asarray(t_hot <= t_cold)).any():
            raise ValueError("t_hot must be greater than t_cold")
        return (t_hot - t_cold) / np.log(t_hot / t_cold)

    @staticmethod
    def celsius_to_kelvin(t_celsius: Union[float, xr.DataArray, np.array]) -> Union[float, xr.DataArray, np.array]:
        if (np.asarray(t_celsius) > 200).any():
            raise ValueError("t_celsius > 200. Are you sure you are using the right units?")
        return t_celsius + 273.15

    @staticmethod
    def logarithmic_mean(t_hot: Union[float, xr.DataArray, np.ndarray], t_cold: Union[float, xr.DataArray, np.ndarray]) -> Union[float, xr.DataArray, np.ndarray]:
        if (np.asarray(t_hot <= t_cold)).any():
            raise ValueError("t_hot must be greater than t_cold")
        return (t_hot - t_cold) / np.log(t_hot / t_cold)